Given that 3.5 haemoherraged fans and 4e seemed to stop the rot this is questionable.
If we look at Google Trends in the four years starting six months after the release of D&D 3.5 more than half the D&D searches dropped off. In the
five years starting six months after the release of 4e, the drop off was around 25%.
To steal an adage from Richard Feynman - The first principle is to not fool yourself, and you are the easiest person to fool.
Yes, that's what was happening with Google searches.
The question, of course, is
why? You rather blithely imply that these numbers reflect the rate at which people were outright leaving the game, and, by extension, the drop off was due to some characteristic(s) of the games in question.
Isn't that just a big hairy guess, though? How do you defend that as anything other than having data say what you want it to say?
I mean, how do you know it is of people leaving? Why couldn't it just as easily be read as saying that searches dropped off because people had gotten caught up on each new edition's changes, and simply didn't *need* to search any more? It would then make sense that searches for 3.5e would drop off much faster, because it was less of a departure from the previous ruleset, and thus easier to catch up on. They're still playing, they just didn't need the web to tell them as much.
Or, more specifically, they don't need *Google* to tell them as much - if they find sources they bookmark, they no longer need Google. So, maybe 3.5e had more *bookmarkable* sources that folks could reliably return to for information, leading to a faster drop-off of searches? Maybe it is about the websites, not about the games themselves.
Or, we could take this as a critique upon the games - in general, 4e requires more external input for people to understand than 3.5e does!
Or, we could note that 4e was released in 2008. Completely unrelated to the release, the years following it were economically disastrous for many people - in tough times, folks may tend turn to inexpensive forms of entertainment - maybe there were more 4e searches because cruising the internet for stuff about RPGs is a heck of a lot cheaper than most other ways to spend your time and keep your mind engaged, and 4e had the "benefit" of being the new kid on the block at the time.
Any of those, of course, would be just more big hairy guesses. All are consistent with the data, but only our *feelings*, not data, tell us one is more probably true than the other.
The point being that plausible story doesn't equal truth.